This environment provides a fully sharded mongo environment using docker-compose and local storage.
The MongoDB environment consists of the following docker containers
- mongosrs(1-3)n(1-3): Mongod data server with three replica sets containing 3 nodes each (6 containers)
- mongocfg(1-3): Stores metadata for sharded data distribution (3 containers)
- mongos(1-2): Mongo routing service to connect to the cluster through (1 container)
sudo apt install docker.io
sudo apt install python-pip
sudo pip install docker-compose
git clone git@github.com:reisdebora/mongodatamodels.git
cd mongodatamodels
This will pull all the images from Docker hub and run all the containers. Please note that you will need docker-compose 1.4.2 or better for this to work due to circular references between cluster members. You will need to run the following once only to initialize all replica sets and shard data across them
./mongoshard.sh init
You should now be able connect to mongos1 and the new sharded cluster from the mongos container itself using the mongo shell to connect to the running mongos process
docker exec -it mongos1 mongo --port 21017
mongoshard.sh {start|stop|restart|reset|init|status}
Data is stored at ./data/ and is excluded from version control. Data will be persistent between container runs. To remove all data ./mongoshard.sh reset